High‐resolution velocity‐azimuth joint estimation for random‐time‐division‐multiplexing multiple‐input‐multiple‐output automotive radar using matrix completion. Issue 10 (31st May 2021)

• Record Type: Journal Article
• Authors: Hu, Xueyao; Zhang, Liang; Long, Jiamin; Liang, Can; Liu, Jianhu; Wang, Yanhua
• Is Part Of: IET radar, sonar & navigation. Volume 15:Issue 10(2021)
• Page Start: 1281
• Page End: 1296
• Publication Date: 2021-05-31
• Subjects: Signal processing -- Periodicals; Radar -- Periodicals; Sonar -- Periodicals; Electronics in navigation -- Periodicals; Navigation -- Periodicals; 621.3848
• DOI: 10.1049/rsn2.12110
• Languages: English
• ISSNs: 1751-8784

Abstract

Abstract: Due to its low hardware complexity, small volume and simple structure, time‐division‐multiplexing multiple‐input‐multiple‐output (TDM MIMO) radar has been widely applied in automotive applications. The transmitting antennas of TDM MIMO automotive radar are usually switched according to a sequential‐TDM pattern. However, moving targets can introduce a motion‐induced phase into the sequential‐TDM pattern, which is coupled to the spatial phase, resulting in errors in velocity and azimuth estimation. Herein, a random‐TDM pattern with the matrix completion (MC)‐based estimation method is proposed to address these issues. In the proposed method, the random‐TDM pattern, which means randomly activating the transmitting antenna instead of sequentially, can decouple the linear temporal‐spatial coupling phase, avoiding the coupling problem in velocity and azimuth estimation. Then, by reshaping the echo data into a matrix form of sparse sampling, the inexact augmented Lagrange multiplier algorithm is adopted to recover this matrix, solving the underdetermined estimation problem caused by sparse sampling. Finally, the high‐accuracy velocity and azimuth can be jointly estimated by applying the estimation of signal parameters via rotational invariance technique algorithm to the recovered full sampling data. Moreover, compared with the compressed sensing‐based method, the proposed method overcomes the grid mismatch problem. The results of comparative simulations and real‐data experiments demonstrate the feasibility of the proposed method.